From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugzilla@gcc.gnu.org>
Received: by sourceware.org (Postfix, from userid 48)
	id 677463858D32; Sat,  3 Jun 2023 08:58:28 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 677463858D32
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1685782708;
	bh=7Mga5+p+9Jfc7WhYaUiDyVZvi7z3EAVb9Mg8PEraubc=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=eyTT5H9CuC+MftVC4NwIXSirIXK454w4nPiA91ITPSI8KitbfflznVa5jfpBET2w5
	 EJmszWODOLCpiZQ1QARJ10UhgE6iaEP5CilbnN7LblJwv2g5514tKEK6D/P5AjGl01
	 jBPMov0ZiDEysbMZmYZnIZQDKWeJWOw9HNnnHwig=
From: "olegendo at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug rtl-optimization/58517] ifcvt (after combine) puts ccreg
 clobbering insn between ccset insn and ccreg use
Date: Sat, 03 Jun 2023 08:58:28 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: rtl-optimization
X-Bugzilla-Version: 4.9.0
X-Bugzilla-Keywords: wrong-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: olegendo at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-58517-4-CfOj73baeq@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-58517-4@http.gcc.gnu.org/bugzilla/>
References: <bug-58517-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
List-Id: <gcc-bugs.sourceware.org>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D58517

--- Comment #13 from Oleg Endo <olegendo at gcc dot gnu.org> ---
(In reply to Andrew Pinski from comment #12)
> Created attachment 55239 [details]
> Patch which does work on this
>=20
> Though, I need to double to make sure it works for other cases still.
> sh is the case where we have a non-CC mode register for the flags which is
> why this was harder than other targets.

Thanks for looking at this (after 10 years .. wow, time flies).


> diff --git a/gcc/config/sh/sh.md b/gcc/config/sh/sh.md
> index 4622dba0121..ae1a0f7eb7c 100644
> --- a/gcc/config/sh/sh.md
> +++ b/gcc/config/sh/sh.md
> @@ -1510,7 +1510,7 @@ (define_expand "movsicc"
>    rtx op0 =3D XEXP (operands[1], 0);
>    rtx op1 =3D XEXP (operands[1], 1);
>
> -  if (! currently_expanding_to_rtl)
> +  if (0 && ! currently_expanding_to_rtl)
>      FAIL;
>
>   switch (code)
> diff --git a/gcc/config/sh/t-sh b/gcc/config/sh/t-sh
> index 1cc8c39d2a8..8ffcc288cf3 100644
> --- a/gcc/config/sh/t-sh
> +++ b/gcc/config/sh/t-sh
> @@ -89,8 +89,8 @@ MULTILIB_OSDIRNAMES =3D \
>  	m4a=3D!m4a $(OTHER_ENDIAN)/m4a=3D!$(OTHER_ENDIAN)/m4a \
>  	m4al=3D!m4al $(OTHER_ENDIAN)/m4al=3D!$(OTHER_ENDIAN)/m4al
>=20=20
> -$(out_object_file): gt-sh.h
> -gt-sh.h : s-gtype ; @true
> +#$(out_object_file): gt-sh.h
> +#gt-sh.h : s-gtype ; @true
>=20=20
>  # Local Variables:
>  # mode: Makefile

Are these hunks intentional?=